以太坊研究员Drake呼吁加密行业进入“掩体模式”防御AI破解密码
Initiate ‘bunker mode’ to protect crypto from AI, says Justin Drake
Senior Ethereum Foundation researcher Justin Drake, who has spent months warning about the supposedly imminent cybersecurity threat of quantum computing, has escalated his rhetoric with a prediction that AI will “break” elliptic curve cryptography within “months.”
以太坊基金会高级研究员贾斯汀·德雷克（Justin Drake）花了数月时间警告所谓迫在眉睫的量子计算网络安全威胁，如今他升级了言论，预测人工智能将在“几个月内”“破解”椭圆曲线密码学。
Warning the threat could arrive even faster than quantum computers, he’s calling for a mass migration of most crypto assets into new wallets suited for deep cold storage.
他警告说，这一威胁的到来速度甚至可能比量子计算机还要快，并呼吁将大多数加密资产大规模迁移至适合深度冷存储的新钱包。
He wants the whole industry, starting with the largest entities, to hide their money from an impending AI breakthrough that he views as essentially inevitable. He’s implored everyone to create and store most digital tokens in new “bunker” wallets that have never signed a transaction.
他希望整个行业——从最大的实体开始——将资金隐藏起来，以躲避一场他认为本质上不可避免的人工智能突破。他恳请所有人创建并使用从未签署过交易的全新“地堡”钱包来存储大部分数字代币。
These more private wallets with less information revealed on public ledgers will, in Drake’s view, help digital asset owners stay safe during the first wave of attacks.
在德雷克看来，这些在公共账本上披露信息较少的更私密钱包，将帮助数字资产所有者在第一波攻击中保持安全。
The post is, in essence, a description of a well-known best practice by BTC holders: never re-using addresses.
这篇帖子本质上是在描述比特币持有者众所周知的一项最佳实践：绝不重复使用地址。
Satoshi Nakamoto even wrote in October 2008: “A new key pair should be used for each transaction to keep them from being linked to a common owner.”
中本聪甚至在2008年10月写道：“每笔交易都应使用新的密钥对，以防止它们被链接到同一个所有者。”
Explaining this concept as though it was novel, Drake thanked Satoshi’s large and highly public wallets that will ostensibly shield smaller investors from the first wave of attacks.
德雷克将这个概念解释得仿佛它是全新的，并感谢了中本聪那些规模庞大且备受公众关注的钱包，称其表面上将为小型投资者抵御第一波攻击提供保护。
Make no mistake, however, Drake believes the AI is coming for everyone soon.
然而，不要误解，德雷克相信人工智能很快就要波及每个人。
I don't recommend anyone scramble to move their funds to new wallets today. But we should take the risks to cryptography from AI-accelerated math seriously, and minimize our exposure to not just quantum-vulnerable cryptography, but also potentially AI-vulnerable cryptography.…
我不建议任何人今天匆忙将资金转移到新钱包。但我们应该认真对待人工智能加速数学带来的密码学风险，并尽量减少我们的暴露面，不仅针对易受量子攻击的密码学，还包括可能易受人攻击的密码学……
— vitalik.eth (@VitalikButerin) October 7, 2026
—— vitalik.eth (@VitalikButerin) 2026年10月7日
Crypto should enter bunkers and cower from the AI
加密货币应进入地堡，躲藏于人工智能之下
With yesterday’s viral X post, Drake asked the industry to commence “bunker mode,” a mass migration to fresh addresses, in a post that drew millions of views.
随着昨天在X平台上病毒式传播的帖子，德雷克在一篇获得数百万次观看的帖子中，要求行业启动“地堡模式”，即大规模迁移至全新地址。
He implores the largest entities and their billionaire operators to get their assets into bunkers first.
他恳请最大的实体及其亿万富翁经营者率先将资产转入地堡。
Drake even named four of the largest custodians by name: Changpeng Zhao’s Binance, Vlad Tenev’s Robinhood, Babak Zanjani’s Bitbank, and Giancarlo Devasini’s Bitfinex and Tether.
德雷克甚至点名指出了四大最大托管机构：赵长鹏的币安（Binance）、弗拉德·特涅夫的罗宾汉（Robinhood）、巴巴克·赞贾尼的Bitbank，以及詹卡洛·德瓦西尼的Bitfinex和Tether。
Four companies with billionaire leadership will guide the industry to safety.
这四家由亿万富翁领导的公司将引领行业走向安全。
A wallet that has never transacted, such as a “bunker” BTC wallet in Drake’s view, exposes only a hash of its public key on-chain. This privacy contrasts with a wallet which has transacted and published its raw public key within its transaction signatures.
一个从未进行过交易的钱包，例如在德雷克看来属于“地堡”类型的比特币钱包，仅会在链上暴露其公钥的哈希值。这种隐私性与那些已进行交易并在交易签名中发布其原始公钥的钱包形成对比。
Because only the hash of a “bunker” wallet is publicly accessible, a quantum attacker who can break Bitcoin’s Elliptic Curve Digital Signature Algorithm must first hack the public key hash and then hack its associated private key to steal any BTC.
由于“地堡”钱包的哈希值是公开可访问的，能够破解比特币椭圆曲线数字签名算法（ECDSA）的量子攻击者必须先破解该公钥哈希值，然后才能破解其关联的私钥以窃取任何比特币。
This extra step of requiring the AI- or quantum-powered hacker to reverse engineer the public key hash first, before reversing the private key, might protect bunkered savers from the first wave of attacks, Drake predicts.
德雷克预测，这一额外步骤——要求借助人工智能或量子技术的黑客先逆向工程公钥哈希值，再逆向私钥——可能会保护处于“地堡”状态的储蓄者免受第一波攻击。
A rational AI or quantum hacker might decide to start with the easiest targets: wallets with exposed public keys, such as Satoshi’s wallets.
理性的AI或量子黑客可能会决定从最简单的目标开始：公钥暴露的钱包，例如中本聪的钱包。
Bitcoin’s quantum deadline just moved up
比特币的量子截止日期提前了
Read more: Cloudflare’s 2029 quantum sprint raises Bitcoin alarm bells
阅读更多：Cloudflare 2029年量子冲刺引发比特币警报
Crypto bunker life
加密货币地堡生活
In Drake’s worldview of bunker life, never-transacting wallets function like deep cold storage, never transacting except in the event of emergency.
在德雷克的“地堡生活”世界观中，从未进行过交易的钱包类似于深度冷存储，除非发生紧急情况，否则从不进行交易。
Any transaction must immediately accompany a full sweep of funds into another, new, never-transacted wallet — ideally within the same block of transactions.
任何交易都必须立即伴随将资金全部转移到另一个新的、从未交易过的钱包中——理想情况下是在同一个交易区块内完成。
He admits plainly, “Exiting bunker mode safely will require post-AI cryptography” that doesn’t currently exist in consensus for blockchains like Bitcoin.
他坦率承认：“安全退出地堡模式需要后AI时代的密码学”，而目前像比特币这样的区块链共识中并不存在此类技术。
In March, Drake co-authored Google Quantum AI’s well-publicized paper suggesting that breaking the Elliptic Curve Digital Signature Algorithm (ECDSA) could occur within 1,200 logical qubits, a 20-fold reduction in the prior understanding of physical hardware requirements for that quantum breakthrough.
今年三月，德雷克与Google Quantum AI共同撰写了一篇广受关注的论文，指出破解椭圆曲线数字签名算法（ECDSA）可能仅需1,200个逻辑量子比特，这将此前对该量子突破所需的物理硬件要求的认知降低了20倍。
In June, Drake forecasted Q-day, i.e. the day quantum computers crack major forms of modern cryptography, at “10% by 2030” and “50% by 2032.”
六月，德雷克预测Q日（即量子计算机破解现代主要加密形式的日子）的概率为“2030年达到10%”和“2032年达到50%”。
Drake’s new fear is non-quantum AI computers. On Tuesday, OpenAI published a trove of AI-generated mathematical results.
德雷克的新担忧是非量子AI计算机。周二，OpenAI发布了一批由AI生成的数学成果。
He interprets the batch as a sign that mathematical superintelligence has arrived. He warned, “In my opinion it is now reasonable to brace for the possibility that ECDSA breaks before Qday, in the worst case in months not years.”
他将这批成果解读为数学超级智能已经到来的信号。他警告称：“在我看来，现在合理预期ECDSA可能在Q日之前就被破解，在最坏的情况下，只需数月而非数年。”
Got a tip? Send us an email securely via Protos Leaks. For more informed news and investigations, follow us on X, Bluesky, and Google News, or subscribe to our YouTube channel.
有线索？请通过Protos Leaks向我们发送加密邮件。如需获取更多深入的新闻报道和调查，请关注我们在X、Bluesky和Google News上的账号，或订阅我们的YouTube频道。
更进一步：量化金融体系
看懂新闻只是起点——沿量化金融路径，把它变成能交付的工程能力